What authorizations reveal about healthcare’s tech ‘arms race’ | Viewpoint

Last year, I needed an MRI for a torn ligament. Like most patients, I expected to wait several days to obtain an authorization for imaging. Unlike most patients, I knew that workflow between my provider and payer would allow for a near-real time determination.

I asked if the front desk staff could try submitting the authorization request while I stood there. They hesitated, but I insisted. Seconds later, the request was approved, and the scan was scheduled before I left the building.

As a healthcare technology executive, I knew my payer had implemented tools capable of evaluating requests against medical policy criteria in real-time. Most patients would have no reason to know such systems exist, much less when or how to ask for them. This wasn’t a loophole; it was one of the rare cases where the system worked as intended. It offered a glimpse of what’s possible when medical policy, clinical data, and workflow infrastructure are aligned.

How authorizations became a systemic liability

For most healthcare stakeholders, providers, payers, and patients alike, authorizations are onerous; a reflection of a system that has grown increasingly complex, fragmented, and adversarial over time.

In her book Thinking in Systems, Donella Meadows writes: “Every system is perfectly designed to get the results it gets.” With utilization management, each stakeholder acts rationally within their siloed domain. Payers tighten controls to manage medical cost and risk. Providers staff up to manage growing administrative demands. Technology vendors optimize isolated transactions. Regulators introduce new rules to drive transparency.

Authorizations originated from a need to ensure appropriate and cost-effective utilization of healthcare resources. However, what started as a clinical safeguard became something else entirely when introduced into a system characterized by fragmented infrastructure, regulatory overload, and competing business models. Authorizations haven’t failed because its fundamental goals were flawed; it failed because its implementation collided with the messy realities of a fragmented multi-stakeholder ecosystem.

Solving authorizations is hard

Authorizations are genuinely hard to solve because the process exists at the intersection of clinical judgment, contractual nuance, regulatory compliance, and fragmented infrastructure and workflows. There is no single owner, no universal definition of “medical necessity,” and no one-size-fits-all solution.

Millions of authorization transactions are processed annually in the U.S., with more than 50 million for Medicare Advantage alone. Each requires some combination of clinical review, data submission, and payer-provider coordination. At the same time, authorization requirements vary widely depending on the service type, and the patchwork of state-level regulations makes nationwide uniformity nearly impossible.

The result is a process that consumes massive amounts of clinical and administrative time without reliably improving care for patients.

Gaps in payer-provider communication

Much of this dysfunction stems from fundamental misalignment between payers and providers. Payers’ goals are to contain costs and prevent member harm by avoiding unnecessary services. Providers’ goals are to deliver care to patients efficiently and be reimbursed appropriately. These motives are not inherently incompatible, but without trust, transparency, and shared accountability, they frequently become adversarial.

Payers often do not transparently disclose the specific clinical criteria used in decision-making, leaving providers to guess what documentation is needed or if an authorization is even required. Criteria may vary across lines of business or delegated vendors, further compounding confusion.

Providers submit requests with variable levels of supporting information, often unsure of what will trigger an approval or a delay. Each side views the other as the source of inefficiency. And without shared data or functioning feedback loops, neither has a complete view of what’s working or what’s failing.

Infrastructure system and delegated entities

Technology exacerbates the problem. Many payer and provider systems are built for compliance, not coordination. The lack of integration between the two leads to rework, phone calls, faxes, appeals, and escalating frustration on both ends.

Adding to the complexity, many health plans outsource portions of the authorization process to delegated entities, utilization management vendors, pharmacy benefit managers, or third-party administrators. While intended to increase scalability, these arrangements often limit payer visibility into decision criteria and turnaround times, further eroding provider trust and complicating resolution paths.

What a better system looks like

There is no shortage of solutions promising to fix authorizations. But most fall into a familiar trap: automating existing workflows without redesigning the system around shared goals. When we look at high-functioning prior authorization models, a few critical principles emerge:

Start with clinical data. Instead of re-keying patient information into a portal, modern systems pull relevant, structured clinical data directly from the electronic health record. This reduces staff burden, improves accuracy, and enables near-instant decisions for many outpatient services.

Make decisions transparently. Rather than a binary yes/no output, advanced platforms show how requests align—or don’t—with codified policy. This transparency reduces appeals, builds trust, and provides a defensible audit trail for both payers and providers.

Unify workflows. Providers are often forced to navigate multiple systems depending on benefit type or delegation. In improved models, all outpatient authorizations enter through a single, bi-directional channel—regardless of who makes the final decision.

Build interoperability into the foundation. With CMS-0057-F accelerating prior authorization modernization, health plans must update their infrastructure. Leading solutions treat interoperability not as a compliance requirement, but as a core design principle that supports care coordination and long-term scalability.

Support clinical review, don’t sidestep it. Some cases will always require human judgment. In high-performing systems, clinical reviewers are presented with structured, policy-aligned case summaries that accelerate decision-making and reduce variability, without displacing clinical expertise.

Use AI to support, not decide. Emerging AI tools can help accelerate determinations, surface relevant clinical context, and reduce variability in decision-making. AI should be used to inform decisions, not to make or deny them. Speed cannot come at the expense of trust. Every output must be transparent, auditable, aligned with evidence-based policy, and ultimately reviewable by a qualified clinical professional.

These aren’t hypothetical improvements. Organizations that have implemented these principles are already seeing faster determinations, lower appeal rates, and improved provider satisfaction. For these organizations, true leverage lies in how decisions are governed and aligned across institutions, not just how fast they move.

A strategic imperative for healthcare leaders

The tools to fix authorizations are here. But the system will not change until executive leaders treat authorizations not as a point solution, but as a reflection of broader organizational design.

That means:

• Aligning incentives across departments, so that success reflects patient impact, not just internal throughput;
• Creating shared governance structures that bring clinical, operational, and technical leaders together;
• Investing in interoperable infrastructure that supports compliance, modernization, and ecosystem collaboration;
• Redefining success metrics to include time to care, appropriateness and consistency of decisions, and measurable reductions in burden

Most health plans and provider organizations already have mandates and strategic imperatives to improve authorizations. What’s often missing is a coherent execution framework, a way to align governance, technology, and measurement across silos. Without that structure, well-meaning initiatives stall in operational limbo.

Leaders who wait for perfect alignment will inherit even deeper dysfunction. Leaders who act now will shape what’s next. The question is whether leadership will choose to redesign, not just react.

Until we shift from a mindset of control and containment to one of alignment and enablement, we’ll keep refining a system no one wants to live with. What we choose to build today will determine whether the next generation inherits a system of care or a system of unnecessary complexity.

Russ Thomas is the CEO of Availity.